Hazard Identification Related to the Presence of Vibrio spp., Biogenic Amines, and Indole-Producing Bacteria in a Non-Filter Feeding Marine Gastropod (Tritia mutabilis) Commercialized on the Italian Market

Tritia mutabilis is a carrion-feeder edible marine gastropod with an open circulatory system. Therefore, biological, and chemical contaminants associated with the feed can reach all body tissues. The aim of the present study was to investigate the possible association of these characteristics with some food safety hazards. Vibrio spp. load, and the prevalence of pathogenic V. parahaemolyticus, V. vulnificus, and V. cholerae, were investigated. Moreover, biogenic amines (BAs) and indole-producing bacteria (IPB), markers of seafood decomposition, were quantified for the first time in an edible carrion-feeder. Overall, 49 batches were analyzed (38 from retail, and 11 from primary production). The Vibrio spp. load resulted of 5.64 ± 0.69 log10 CFU g−1 at retail, and 5.27 ± 0.74 at harvest but all batches resulted negative for pathogenic Vibrio. Histamine, putrescine, cadaverine, and tyramine were detected both at harvest and at the retail level. Their sum (BAs Index) showed a mean value of 50.45 and 65.83 mg Kg−1 in batches at harvest and at retail, respectively. IPB were detected at harvest and upon refrigeration for three days (T1–T3). The mean load resulted in 2.52 ± 0.85 log10 MPN g−1 at T0, 3.31 ± 1.23 at T3 in batches immediately refrigerated, and 3.22 ± 1.18 at T3 in batches previously immersed in clean seawater. Our results contribute to identifying food-borne hazards for T. mutabilis that may be related to the retention of biogenic amines and indole-producing bacteria due to carrion feeding.


Introduction
Fishing of marine gastropod Tritia mutabilis is by far the most important activity carried out by artisanal fisheries in the Central and Northern Adriatic Sea, yielding from 2000 to 3000 tons of landings each year [1,2].
According to diet, gastropods can be herbivorous, carnivorous, and omnivorous, and their feeding strategies are also very diversified including grazing, suspension-feeding, predation, and even parasitism [3]. T. mutabilis is a scavenger gastropod, feeding on carrion in the adult stage [2,4].
Gastropods, together with bivalve mollusks, possess an open circulatory system, characterized by irregular spaces or sinuses within the tissues between the distributing and collecting blood vessels [5]. The system is considered open because the blood (hemolymph) empties from a contractile heart and major supply vessels into the body cavity (hemocoel), where it directly bathes the organs [6]. Therefore, contaminants including bacteria, viruses, and chemicals circulating after feeding can easily reach all body tissues.
The genus Vibrio comprises the major culturable bacteria in marine and estuarine environments [7], and with 130 confirmed species, it represents one of the most diverse marine bacterial genera, geographically spread all over the world, but more commonly occurring in warmer waters [8]. The genus Vibrio includes several food-borne pathogens that cause a spectrum of clinical conditions, that can be grouped into three major syndromes: gastroenteritis, wound infection and septicemia [9]. Pathogenic vibrios have emerged as new threats since the 1970s [10], causing an increased incidence of human infections worldwide [11][12][13], and the consumption of raw or undercooked bivalve mollusks is considered the primary source particularly oyster, clam, and mussels, because of their filterfeeding habit that allows concentrating particles associated with various contaminants from the surrounding waters [14][15][16]. Unfortunately, marine gastropods, unlike bivalves, are poorly investigated in terms of food safety.
Nevertheless, the European Union (EU) legislation does not include pathogenic vibrios, biogenic amines and other products of the decomposition among the safety criteria for gastropods, regarded as equivalent to bivalve mollusks, echinoderms, and tunicates [17]. To our best knowledge, no outbreaks of food poisoning caused by T. mutabilis have ever been reported.
Appropriate cooking treatment is effective to inactivate microorganisms associated with seafood, but it is considered little or nothing effective on their toxic products, as documented for many marine biotoxins [18,19], biogenic amines (BAs) [20], and indole, whose production is fostered by high pH and temperature [21].
BAs are organic bases with aliphatic, aromatic, and heterocyclic structures produced mainly by bacteria [20]. During the decomposition of seafood, various amounts of BAs are usually produced, in particular, the Gram-negative bacteria are considered the predominant amine-forming bacteria in fish, cephalopods, and shellfish [22], including Vibrio spp. [23]. BAs are formed mainly by decarboxylation of specific free amino acids or by amination and transamination of ketones and aldehydes [23,24]. Decarboxylation of histidine, lysine, ornithine, and tyrosine results in the production of histamine, cadaverine, putrescine, and tyramine respectively [25], which are commonly used as chemical indicators of fish decomposition [22]. BAs in food constitutes a potential public health concern due to their physiological and toxicological effects, such as histamine intoxication, migraine, and food intolerance crises [26]. Except for histamine in fishery products no threshold is established for these biomolecules in EU legislation. On the other hand, despite a widely reported association between histamine and scombroid food poisoning, histamine alone appears to be insufficient to cause food toxicity, as other BAs, i.e., putrescine and cadaverine, may potentiate histamine toxicity [27].
Indole is widespread in nature because various species of bacteria can produce a considerable amount of this heterocyclic aromatic amine [28], via the action of the enzyme tryptophanase, expressed in many Gram-negative, as well as Gram-positive bacteria [29]. It should be noted that members of the genus Vibrio are nearly all indole positive [30]. Indole may serve as a chemical indicator for the evaluation of incipient spoilage of crustaceans and other seafood products [31] and has long been regarded as a typical N-heterocyclic aromatic pollutant due to its toxicity and potential mutagenicity causing animal hemolysis, hemoglobinuria nephrosis, temporary skin irritation and tumor formation [21]. Normally, indole is metabolized to indoxyl sulfate in the liver and then excreted into the urine, but if the dietary intake exceeds the renal clearance, the serum levels of indoxyl sulfate increase with toxicological effects [2].
Official controls on non-filter feeding marine gastropods harvested from non-classified production areas are considered not yet fully implemented by the EU Member States [32], in any case, several food-borne hazards are not considered by the Legislation in force. Among them, pathogenic vibrios, BAS and indole.
The aim of the present study was to ascertain possible food-borne hazards related to the accumulation of bacteria and chemical products of the decomposition in T. mutabilis from Emilia-Romagna Region, Northern Italy. Batches at harvest and retail were considered under different conditions for viability of the batches, Vibrio spp. load, and the prevalence of pathogenic species, namely V. cholerae, V. parahaemolyticus, and V. vulnificus. To ascertain the possible accumulation of toxic catabolites assumed from the carrion, Biogenic Amines (BAs) and Biogenic Amines Index (BAI) were investigated at harvest and retail, whereas Indole Producing Bacteria (IPB) load was investigated at harvest and under refrigerated storage.

Materials and Methods
The study area is part of the FAO zone 37.2.1, GSA (Geographical Subarea) 17, along the Emilia-Romagna coastline, Adriatic Sea, Northern Italy ( Figure 1). T. mutabilis is not available year-round, because the fishing season is from the early autumn to the late spring. Therefore, the study required several surveys carried out between October 2017 and February 2021.
The research program was divided into three Work Packages (WPs): (WP1) evaluation of batches viability, Vibrio spp. load and the prevalence of pathogenic vibrios at retail; (WP2) evaluation of batches viability, Vibrio spp. load, and the prevalence of pathogenic vibrios and BAs content at retail and at harvest; (WP3) evaluation of batches viability, Vibrio spp. load, prevalence of pathogenic vibrios and IPB load at harvest and during refrigeration at 4 • C with and without prior treatment of immersion in clean seawater in a Recirculating Aquaculture System (RAS). The treatment is normally intended for bivalve mollusks depuration (to reduce E. coli contamination) and conditioning (to remove sand, mud or slime, to improve organoleptic qualities and to ensure good vitality).
The batches from primary production were obtained directly from a fisherman (at landing), and those at retail from different fish shops. All batches were transported immediately to the laboratory in isothermal boxes, stored at 6-8 • C to avoid Vibrio spp. entering a viable but non-culturable state, and processed within 24 h.

Viability of the Batches (WPs 1, 2, and 3)
According to the EU legislation [17], "fresh" marine gastropods must be sold alive as bivalve mollusks, to which they are assimilated. They must show characteristics of freshness and viability, must be responsive to the touch by closing their open valves, and must contain normal amounts of intravalvular liquid. This approach is tailored to the specific features of bivalves, and obviously, it is unsuitable for gastropods, because they lack two valves and obviously, intravalvular liquid. The viability of the batches of T. mutabilis was ascertained following an internal protocol, by spreading table salt on the animals, as NaCl results in an aversive stimulus promoting extroflexion of the foot and mucus production in viable animals [33]. For each batch, the viability was expressed as the percentage of living animals on the whole batch.
The sample units (10-20 individuals) were prepared according to a slight modification of the ISO 6887-3:2017/Amd.1:2020 method. Briefly, gastropods were rinsed with sterilized seawater, then their shells were cut aseptically to obtain 10 g of the whole body. Each unit was added to 90 mL of saline solution NaCl 3% and homogenized utilizing a rotary blender at medium speed. From this first dilution, further ten-fold serial dilutions were prepared with saline solution NaCl 3%.
From each homogenate and its ten-fold serial dilutions, 100 µL were spread plated on the agar plates and incubated at 20 • C. The results were expressed as log 10 Colony Forming Units (CFU) g −1 .
V. parahaemolyticus, V. cholerae and V. vulnificus were investigated utilizing the homogenate of each sample unit, from which 100 µL were spread plated on agar plate of CHROMagar™ Vibrio (CAV) (PBI International, Milan, Italy), then incubated at 37 • C for 24 h. Suspected colonies (V. parahaemolyticus mauve, V. vulnificus/V. cholerae green-blue to turquoise blue, V. alginolyticus colorless) were submitted to biochemical screening and genotyped by Polymerase Chain Reaction (PCR). Briefly, after a biochemical screening to ascertain the characteristics of the genus (pleomorphic Gram-negative rods, oxidase-positive, able to reduce nitrate, glucose-fermenting, sensitive to the vibriostatic O129/150 µg), suspected strains were genotyped by PCR as previously described [35] to test their respective species-specific and pathogenic gene markers. For V. parahaemolyticus toxR, tdh and trh; for V. cholerae toxR, ctx and stn/sto; for V. vulnificus only the species-specific gene marker vvhA, because, according to the Food and Agricultural Organization of the United Nations and the World Health Organization [36], all V. vulnificus strains may be prudently considered virulent. The oligonucleotide primers and the PCR conditions are reported in Supplementary Materials Table S1.

Biogenic Amines (WP2)
BAs were quantified by an HPLC method employing a UV detector after derivatization by Dansyl chloride. Analyses were performed by an external service (FoodMicroTeam s.r.l., Academic Spin-off of the University of Florence, Italy), and samples were supplied frozen (−20 • C). Considering the results for each batch, the sum of histamine, putrescine, cadaverine, and tyramine quantities were calculated and expressed as BAI (Biogenic Amines Index) according to Veciana-Nogués et al. [37].

Abundance of IPB (WP3)
IPB load was determined in T. mutabilis by Most Probable Number procedure, indicated by Leitão and Rios [38] as the most reliable approach for the quantitative evaluation of the indole positive microflora. Three series of three tubes containing 5 mL of tryptone broth 1% were prepared. One series was inoculated with 0.5 mL of the homogenate (dilution 1:10), and the remaining two with 0.5 mL of two further ten-fold serial dilutions each. After incubation at 20 • C for 3-5 days, Kovac's reagent was added to each tube. The reagent builds with indole a cherry-red complex, allowing to confirm the enzymatic activity of tryptophanases and the positivity for IPB.

Statistical Analysis
Statistical analysis was performed on Vibrio spp. load, BAs, BAI, and IPB content of the batches of T. mutabilis. According to the EU legislation [17], bivalve mollusks and gastropods placed on the market must be alive, but a mortality ≤10% is almost always accepted. For this reason, batches were divided into two groups based on the percentage of viability: high viability (≥90%) and low viability (<90%). The one-sample Kolmogorov-Smirnov (K-S) test was performed to assess data distribution. According to K-S test results, either non-parametric Mann-Whitney U test (M-W U) or Student's t-test were used to establish possible differences in Vibrio spp. load, BAs, BAI, and IPB content between the two viability groups or between treatment groups (with or without immersion). Paired samples t-test was used to compare viability, Vibrio spp. load and IPB content between T0 and T3 within treatment groups.
Statistical significance was set at p ≤ 0.05. All statistical analyses were performed using the software SPSS 26.0.0 (IBM SPSS Statistics, Armonk, NY, USA).

Results
Overall, forty-nine batches of T. mutabilis were processed. Results are reported below for each of the three WPs. All batches were negative for V. parahaemolyticus, V. vulnificus, and V. cholerae.

Viability of the Batches, Vibrio spp. Load and BAs Content at RETAIL and at Harvest (WP2)
Overall, 18 batches of T. mutabilis were analyzed ( Table 2). The four batches from primary production showed viability of 100%. Two of the fourteen batches from retail had a viability of 90%, ten < 90% and two 0%. Vibrio spp. mean load resulted in 5.60 ± 0.61 log 10 CFU g −1 in the batches from retail, and 5.30 ± 0.85 log 10 CFU g −1 in the batches at harvest. Histamine, putrescine, cadaverine, and tyramine were detected in all samples and each BAI (their sum) was calculated. All BAI value resulted >15 mg Kg −1 , and 50% (9 out of 18) resulted >50 mg Kg −1 (min-max: 23.2-225.1 mg Kg −1 ). The mean value resulted in 50.45 ± 21.27 Kg −1 in the four batches from the primary production, and 65.83 ± 48.36 mg Kg −1 in the fourteen batches from retail. No statistically significant difference of Vibrio spp. load, histamine, putrescine, cadaverine, tyramine, and BAI quantity were observed between batches from the primary production or from retail.

Viability of the Batches, Vibrio spp. and IPB Load at Harvest and during Refrigeration (WP3)
Seven batches at harvest were utilized to verify any influence of the refrigerated storage at 4 • C for three days (T1, T2, T3) on the viability of the batches, Vibrio spp. load and IPB load. Three batches were immediately refrigerated, while four batches were partly immediately refrigerated and partly refrigerated after 18-24 of immersion in clean seawater in a RAS. Results are reported in Table 3. Overall, at harvest (T0) the mean value of viability resulted in 98.63 ± 3.63% (six batches 100% and one batch 90.4%). The mean value of viability at T3 resulted in 62.43 ± 26.95% in the samples immediately refrigerated and 94.25 ± 7.59% in the samples previously immersed in clean seawater. The mean load of Vibrio spp. resulted in 5.08 ± 0.82 log 10 CFU g −1 at T0, 5.10 ± 0.62 at T3 without immersion, and 5.46 ± 0.12 at T3 with previous immersion.
In batches refrigerated after 18-24 of immersion in clean seawater in a RAS, no significant differences (p > 0.05) were observed in viability, Vibrio spp. load and IPB content between T0 and T3. Similarly, in batches refrigerated without immersion, no significant differences (p > 0.05) were observed in Vibrio spp. load and IPB content between T0 and T3. However, viability at T3 (62.4%) was significantly lower (paired-t: 3.94; p = 0.008) than that observed at T0 (98.6%).

Discussion
The EU legislation assimilates the marine gastropods to the bivalve mollusks in terms of safety criteria for human consumption but allows the harvesting of gastropods that are not filter-feeders in non-classified areas [39,40]. This provision relies on earlier statements expressed in Regulation UE 558/2010 [41]: the risk of accumulation of microorganisms related to fecal contamination may be considered remote in animals that are not filter-feeders; no epidemiological information was reported to link the provisions for classification of production areas with risks for public health associated with marine gastropods which are not filter feeders.
Bivalve mollusks and gastropods have both an open circulatory system [5], and therefore bacteria and other contaminants can be expected in all organs and tissues, independently of the different feeding habits. Despite this biological evidence, the microbiology of marine gastropods has been poorly investigated, except few studies concerning disease outbreaks of Abalone, reporting high bacterial contamination of tissues also in healthy animals [42], and the study of Cheng et al. [43] reporting high contamination of the gastropod Niotha clathrata muscle by Vibrio spp. (6-8 log 10 CFU g −1 ).
A recent study evidenced a Vibrio spp. mean load exceeding 5 log 10 CFU g −1 in the tissues of T. mutabilis and 5.79 log10 CFU g −1 in Bolinus brandaris, another carnivorous gastropod fished, commonly by-catch, in the Adriatic Sea [33].
Accordingly, in the present study Vibrio spp. mean load resulted in >5 log 10 CFU g −1 in all the three WPs. In WP1 resulted in 5.70 ± 0.75 log 10 CFU g −1 in batches from retail (Table 1). In WP2 resulted in 5.60 ± 0.61 log 10 CFU g −1 in the batches from retail, and 5.30 ± 0.85 log 10 CFU g −1 in the batches from primary production ( Table 2). In WP3 resulted in 5.08 ± 0.82 log 10 CFU g −1 in the batches from primary production (T0 = before refrigeration), 5.10 ± 0.62 log 10 CFU g −1 after refrigeration (T3) without previous treatment of immersion, and 5.46 ± 0.12 log 10 CFU g −1 after refrigeration (T3) with previous treatment of immersion (Table 3). It should be noted that these values result higher than those ascertained in R. philippinarum in the same area, showing a mean load of 4.69 ± 0.65 log 10 CFU g −1 in a multi-year retrospective study [44], and 4.82 ± 0.71 log 10 CFU g −1 in a concurrent project (unpublished data).
Notwithstanding the high Vibrio spp. load, all batches of T. mutabilis resulted negative for pathogenic V. parahaemolyticus, V. vulnificus and V. cholerae, whereas batches of R. philippinarum showed 62% of positivity for at least one of the targets in a concurrent project (unpublished data), as proof of their circulation in the area. The discrepancy between the two species is difficult to explain and needs further investigation. It can only be assumed that the microbial environment and the chemical characteristics of the carrion on which T. mutabilis feed could result in an unsuitable environment for V. parahaemolyticus, V. vulnificus and V. cholerae.
The fishing of T. mutabilis is practiced within three miles of the coastline by dedicated traps with dead fishes as baits, predominantly bluefish. These gastropods use their sense of smell to find carrion, therefore spoiled fish is preferred by some fishermen. Other fishermen prefer thawed white fish of acceptable quality to prevent the bitterness of gastropod flesh.
Given the feeding habit of T. mutabilis, chemical products of the decomposition should be considered of concern, particularly BAs and indole. However, to our best knowledge, this is the first study conducted on a marine scavenger gastropod and comparative data are lacking. Histamine, putrescine, cadaverine, and tyramine, were always detected (Table 2), including 14 batches of T. mutabilis from retail and four batches from primary production. Considering the BAI limit of acceptability suggested for tuna (<50 mg Kg −1 ) and for anchovy, (<15-16 mg Kg −1 ) [25], the BAI values registered in T. mutabilis may be considered unquestionably high. All samples had a value >15 mg Kg −1 , and 50% (9 out of 18) had a value >50 mg Kg −1 (minimum 23.2 mg Kg −1 , maximum 225.1 mg Kg −1 ). While not differing significantly, the BAI mean value resulted higher in the batches from retail (65.83 mg Kg −1 ) than in the batches from the primary production (50.45 mg Kg −1 ) suggesting that, after the harvesting, endogenous production of BAs may occur over time, increasing the total amount of these substances in T. mutabilis tissues.
Indole may serve as a chemical indicator for the evaluation of incipient spoilage of crustaceans and other seafood products [31]. A limit of 250 µg Kg −1 was established by the US Food and Drug Administration, and in several countries, to identify shrimps in the first stage of decomposition, but it was demonstrated that also shrimps with a level of 44.6 µg Kg −1 show the rejection conditions [45]. Almost all species of Vibrio spp. can produce indole via tryptophanase activity [30,46], and the activity of tryptophanase may result increased 21-fold at pH 8-9 [47]. Marine carrion is rich medium in a strongly alkaline environment (pH > 8), and it may be expected to contain a significant amount of this substance. As shown in Table 3, IPB were always detected in T. mutabilis batches at harvest (mean load 2.52 ± 0.85 log 10 MPN g −1 ), and with increasing value after three days of refrigeration (3.31 ± 1.23 and 3.22 ± 1.18 log 10 MPN g −1 without and with immersion respectively) suggesting that IPB are predominantly represented by psychrophilic bacteria metabolically active at chilling temperature (4 • C). Considering that fish meat is naturally rich in tryptophan [48] this IPB load in a carrion-feeder is certainly potentially able to produce a high level of indole.
The immersion in clean seawater is provided by EU legislation for bivalve mollusks [17], but not for marine gastropods. Conversely, our results suggest that this treatment may ensure higher viability to T. mutabilis at retail, as the batches immersed in clean seawater after landing showed higher viability after three days of refrigeration (mean value 94.25%) than those immediately refrigerated (mean value 62.43%). It can be assumed that the low percentage of viability registered for 24 batches at retail (only 20.8% with 90-100% of viability) could be due to the lack of treatment of immersion after landing.

Conclusions
The feeding habit of T. mutabilis and its open circulatory system allows a high accumulation of bacteria (Vibrio spp.) and products of the decomposition associated with the carrion on which the animal feed, i.e., BAs and IPB. Notwithstanding the circulation of pathogenic vibrios in the area, namely V. parahaemolyticus, V vulnificus, and V. cholerae (high percentage of positivity in R. philippinarum) all samples resulted negative. To our knowledge, the present study is the first to address various issues of food safety involving T. mutabilis. For this species, the undiscovered safety issues seem to rely on the retention of biogenic amines and indole-producing bacteria because of the feeding on carrion. Obviously, the utilization of spoiled fish as bait should be discouraged, to prevent an excessive intake of products of the decomposition in T. mutabilis. Moreover, the lack of immersion in clean seawater after landing could be at the origin of the low viability of the batches at retail, with possible endogenous production of BAs.
Institutional Review Board Statement: Ethical review and approval were waived for this study, since invertebrate, with the exception of cephalopods, are not included in Directive 2010/63/EU of the European Parliament which regulates this field in the European Union.

Informed Consent Statement: Not applicable.
Data Availability Statement: Not applicable.